Pancreatic cystic lesions: descriptive epidemiology and natural history. A steady increase in incidence of pancreatic cysts has been documented and is related to widespread use of new imaging technologies, mainly CT and MRI. Incidental pancreatic cyst diagnosis ranges from 2.6% to 20% depending on imaging modality and population. The dilemma is that cysts with malignant potential account for ~50% of all cysts and have a d50% 5-year survival as invasive cancers. Yet, it is often unclear which cysts will progress partly due to the poorly understood epidemiology and natural history of these lesions. Further, regardless of malignant potential, current guidelines recommend either watch and wait or surgical resection with decisions based on cyst size, location, symptoms and patient eligibility for surgery. Cancer prevention is the goal of surgery but the significant morbidity and mortality associated with surgical resection cannot be ignored. Thus, continued research that contributes to finer sub-classification and diagnosis of cysts that are or will progress to cancer s needed to prevent unnecessary surgery, characterize high risk populations, improve prognosis and increase knowledge of the understudied natural history of these lesions. A secondary benefit is the potential that this work also will shed light on pancreatogenesis. Our aims are to: 1) Determine cyst lesion incidence in patients e40 years old who received an abdominal CT or MRI at UCSF from 1996-2015; 2) For cyst patients followed at UCSF, abstract prospective clinical and outcome data from medical, pathology and laboratory records to determine factors related to cyst progression/ malignancy and identify archival biospecimens for banking. Link pancreatic cyst patients with SEER cancer registry data to determine cancer outcomes and; 3) Frequency- match controls (no cyst) to cyst cases by 5-year age group, sex, reason for imaging and imaging modality to determine demographic and clinical factors related to pancreatic cyst risk. We expect to identify about 2,000 incident pancreatic cyst patients from review of ~50,000 studies. Descriptive statistics, adjusted multivariable Cox regression including with time-dependent covariates, and unconditional logistic regression analyses will be used to address the study aims. Achievement of our overarching goal, to develop a richly annotated database registry of abdominal imaging-identified (incidental and non-incidental) pancreatic cyst patients at UCSF, will enable further research of understudied factors related to risk and progression of pancreatic cysts and by type. More importantly, as public registries of pancreatic cysts do not exist, our project is well-positioned to build the infrastructure and resources for support of futue prospective cohorts to identify factors to 1) understand biological mechanisms contributing to cyst development; 2) predict progression and malignancy and; 3) better define at-risk groups for closer monitoring or intervention. A comprehensive understanding of pancreatic cyst epidemiology and natural history is of utmost importance to provide the best options for patient care and has the potential to add new data about early pancreatogenesis to inform screening and therapeutic targets.